What Happened
In January 1966, Joseph Weizenbaum published the article describing ELIZA in Communications of the ACM. ELIZA used pattern matching and scripted transformations to simulate conversation, famously showcasing a “therapist-like” dialog style.
Why It Matters
ELIZA became a landmark in human–computer interaction and NLP history, demonstrating both the allure and the risks of anthropomorphizing text-based systems—an issue that reappears in modern chat-based AI products.
Technical Details
ELIZA did not use statistical learning or neural networks; it relied on rule-based text transformations, making it an instructive contrast to today’s data-driven language models.